Materials Science & Engineering A
Microstructural characterization and mechanical properties of nano-scale/sub-micron
TiB-reinforced titanium matrix composites fabricated by laser powder bed fusion
--Manuscript Draft-Manuscript Number:
MSEA-D-21-02529
Article Type:
Research Paper
Keywords:
laser powder bed fusion; Titanium matrix composite; Ti-6Al-4V; TiB; Strength; Wear
resistance
Corresponding Author:
Eskandar Fereiduni
Hamilton, ON CANADA
First Author:
Eskandar Fereiduni
Order of Authors:
Eskandar Fereiduni
Ali Ghasemi
Mohamed Elbestawi
Abstract:
A B 4 C/Ti-6Al-4V(Ti64) composite powder containing a minor B 4 C content
(0.2wt.%) was developed by a novel technique and was subjected to the laser powder
bed fusion (L-PBF) process within a wide range of laser powers and scanning speeds
to fabricate titanium matrix composite (TMC) parts. The relative density measurement
results revealed that almost fully dense TMC parts could be achieved by optimizing
the process parameters . Compared to the Ti64 case, slightly higher values were
required in the TMC system to achieve the highest relative density. Microstructural
characterisations of the TMC parts revealed the formation of large columnar prior β
grains containing in-situ formed nano-scale/sub-micron TiB needles homogenously
dispersed in a martensitic matrix. While having almost the same ductility, the fabricated
TMC parts showed 25, and 8% higher nanohardness, and compressive yield strength,
respectively, and 12% lower wear rate than the Ti64 sample. The improved mechanical
properties of the TMC part were due to the contribution of several factors including the
incorporation of nano-scale/sub-micron TiB reinforcement, the refinement of the
martensite α՛ laths, and the solid solution strengthening effects of carbon atoms. The
contribution of TiB presence and solid solution strengthening was found to be ~70%,
and ~30% in the overall yield strength enhancement of the TMC parts, respectively.
Suggested Reviewers:
Essa Khamis
University of Birmingham School of Mechanical Engineering
k.e.a.essa@bham.ac.uk
Jill Urbanic
University of Windsor
jurbanic@uwindsor.ca
Tushkar Borkar
Cleveland State University
T.Borkar@csuohio.edu
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Cover Letter
Dear Professor Lavernia,
Please find enclosed our manuscript entitled "Microstructural characterization and mechanical
properties of nano-scale/sub-micron TiB-reinforced titanium matrix composites fabricated by laser
powder bed fusion", which we would like to submit for publication in your respected journal, Materials
Science and Engineering: A.
Metal matrix composites (MMCs) are outstanding engineering materials with tailorable properties,
having a huge potential to be used in automotive, aerospace, biomedical, and defense industries.
Nevertheless, their application has been overshadowed by alloys and superalloys over the past few
decades due to the difficulties associated with the processing of MMCs. Laser powder bed fusion
(LPBF) process is one of the most promising additive manufacturing (AM) techniques in terms of
fabricating MMCs with fine features and high dimensional accuracy, which has attracted a great deal
of attention in recent years. Due to the absence of different nozzles for feeding composite constituents,
fabrication of MMCs by the LPBF process requires a composite powder feedstock as the starting
material. Since no commercial powder is currently available for fabricating MMC parts, different
methods have been utilized in recent years to develop composite powder feedstocks for LPBF
applications, among which the ball milling is the most frequently used process.
Although the ball milling provides the attachment of the reinforcing powder constituent to the metallic
powder, it causes the deformation of the metallic powder particles from spherical to non-spherical, and
adversely affects the flowability (as a major characteristic of the powder for LPBF process) of the
obtained composite powder. In this study, a novel technique is proposed to produce 0.2wt.%B4C/Ti64
composite powders for the LPBF process, which benefits from the advantages of the ball milling and
regular mixing methods. The obtained composite powder had a flowability and an apparent packing
density close to those of the monolithic Ti64 powder. A wide range of process parameters were utilized
to fabricate parts out of the composite and monolithic Ti64 powders. The process parameters leading to
the highest densification levels were employed to fabricate compression, and wear test specimens.
Results revealed that almost fully dense TMC parts could be achieved by optimizing the process
parameters. Microstructural characterisations of the TMC parts revealed the formation of large
columnar prior β grains containing in-situ formed nano-scale/sub-micron TiB needles homogenously
dispersed in a martensitic matrix. While having almost the same ductility, the fabricated TMC parts
showed 25, and 8% higher nanohardness, and compressive yield strength, and 12% lower wear rate than
the Ti64 counterpart, respectively. The improved mechanical properties of the TMC part were due to
the contribution of several factors including the incorporation of nano-scale/sub-micron TiB
reinforcement, the refinement of the martensite α՛ laths, and the solid solution strengthening effects of
carbon atoms. While ~70% of this improvement was due to the direct and indirect strengthening effects
caused by the TiB phase, the remaining ~30% was originating from the solid solution strengthening
mechanism.
We confirm that this manuscript has not been published elsewhere and is not under consideration for
publication by another journal. I would highly appreciate it if you could consider this manuscript for
publication in the Materials Science and Engineering: A journal.
Kind regards,
Eskandar Fereiduni
Department of Mechanical Engineering, McMaster University
Hamilton, Ontario, Canada
Manuscript
Click here to view linked References
Microstructural characterization and mechanical properties of nano-scale/submicron TiB-reinforced titanium matrix composites fabricated by laser powder
bed fusion
Eskandar Fereiduni *, Ali Ghasemi *, Mohamed Elbestawi
Department of Mechanical Engineering, McMaster University, Hamilton, ON L8S 4L7, Canada
* Corresponding authors: Eskandar Fereiduni (fereidue@mcmaster.ca), and Ali Ghasemi (ghasemia@mcmaster.ca)
Abstract
A B4C/Ti-6Al-4V(Ti64) composite powder containing a minor B4C content (0.2wt.%) was developed by a
novel technique and was subjected to the laser powder bed fusion (L-PBF) process within a wide range of
laser powers and scanning speeds to fabricate titanium matrix composite (TMC) parts. The relative density
measurement results revealed that almost fully dense TMC parts could be achieved by optimizing the
process parameters. Compared to the Ti64 case, slightly higher 𝐸𝑣 values were required in the TMC system
to achieve the highest relative density. Microstructural characterisations of the TMC parts revealed the
formation of large columnar prior β grains containing in-situ formed nano-scale/sub-micron TiB needles
homogenously dispersed in a martensitic matrix. While having almost the same ductility, the fabricated
TMC parts showed 25, and 8% higher nanohardness, and compressive yield strength, respectively, and 12%
lower wear rate than the Ti64 sample. The improved mechanical properties of the TMC part were due to
the contribution of several factors including the incorporation of nano-scale/sub-micron TiB reinforcement,
the refinement of the martensite α՛ laths, and the solid solution strengthening effects of carbon atoms. The
contribution of TiB presence and solid solution strengthening was found to be ~70%, and ~30% in the
overall yield strength enhancement of the TMC parts, respectively.
Keywords: Laser powder bed fusion; Titanium matrix composite; Ti-6Al-4V; TiB; Strength; Wear
resistance
1. Introduction
Ti-6Al-4V (referred to as Ti64 hereafter) is the most widely used titanium alloy with an (α+β) two phase
microstructure, which makes of more than half the usage of titanium alloys. This alloy offers a good balance
of properties such as high specific strength, adequate stiffness, appropriate high-temperature stability and
resistance, good fatigue behavior as well as outstanding corrosion resistance, making it applicable in a wide
range of industries such as aerospace, petrochemical, and biomedical [1, 2]. Among these industries, the
aerospace sector accounts for >70% of the whole Ti consumption worldwide [3]. Other than these
industries, Ti alloys are rarely used in other sectors due to their relatively high cost [4]. Although having
numerous benefits, the limited wear resistance of Ti alloys is a major concern when high abrasion and
erosion resistance is required [2]. Accordingly, a great deal of attempt has been made in recent decades
toward adding alloying elements or incorporating reinforcing particles (ex-situ and in-situ) to/into Ti alloys
to improve not only their hardness and strength, but also their wear resistance [5].
In the as-cast condition, the microstructure of Ti64 alloy consists of very coarse prior β grains with sizes in
the order of few mm, making it necessary to employ several thermo-mechanical processing steps with the
purpose of breaking these coarse grains down to sub-mm or micron length scales [6]. The melting, thermomechanical processing and final machining stages employed to produce finished components of Ti alloys
have been reported to be very expensive, accounting for approximately 60% of the total cost [7]. A vast
majority of research has been carried out in recent decades to reduce the cost of finished Ti products and
expand their usage to other industrial sectors. For this purpose, it has been tried out to decrease the number
of processing steps of Ti alloys via adding alloying elements which are capable of refining the grain size of
the as-cast microstructure. The refined microstructures will no longer require a number of high-temperature
processing steps, leading to a significant drop in the price of manufactured components. Among a wide
variety of elements acting as grain refiner in Ti alloys, boron (B) has an incredible grain refining effect and
can reduce the as-cast grain size of Ti alloys by an order of magnitude even when existing as minor amounts
(e.g., 0.1wt.%) [8]. Although the reduced cost of Ti alloys obtained by the alloy design eliminates the high
temperature processing steps, Ti alloy components obtained by conventional manufacturing processes are
still expensive. Due to its numerous advantages over traditional manufacturing, additive manufacturing
(AM) has become a direct manufacturing technology with applications across a variety of industries [9, 10].
This unprecedented technology provides fabrication of customized and near-net-shape parts with complex
shapes, fine features, and unique geometries in one shot, making it economically attractive and of a high
interest especially in applications demanding low volume production of relatively expensive materials such
as Ti alloys [11-14]. Accordingly, there is a growing global interest in implementing AM technologies
providing improved material efficiency and lower cost.
Various AM processes have been emerged in recent decades in which the common theme is the
consolidation of the feedstock material in a layer-by-layer manner through the localized melting and
subsequent solidification, sharing similarities to the cast microstructures. Direct energy deposition (DED)
and powder bed fusion (PBF) are the AM categories most frequently used to fabricate metallic parts. The
microstructures of as-deposited Ti64 alloy parts obtained by the DED processes are featured by large
columnar prior β grains extended over multiple layers, surrounded by grain boundary α extended along the
prior β grain boundaries [15, 16]. The grain boundary α is known as the major impediment to the ductility
by providing a continuous pathway for the crack propagation [17, 18].
Recently, there has been a great interest toward engineering the microstructure via controlling the process
parameters during the AM process. Although this strategy has been found to be successful in tailoring the
microstructure within a single component in some alloys (e.g. Inconel 718) [19], the low thermal gradients
needed for prior β grains of Ti64 alloy to form in an equiaxed morphology is not feasible in the processing
space of the DED and PBF techniques [20]. Also, the relatively narrow range of optimum processing
window poses a great challenge to the microstructure tailoring through the engineering of process variables.
Therefore, researchers have been seeking alternative ways of grain refinement in the AM-fabricated Ti64
alloy. Successful attempts have been made lately to refine the size of prior β grains in AM-fabricated Ti
alloys via trace addition of B [16, 20], LaB6 [21], Si [22], and Be [23]. Due to its significantly high growth
restriction factor [24, 25], B has attracted a considerable attention for this purpose. In the DED-fabricated
B-modified Ti64 alloys, B has been shown to not only refine both the prior β and α grain size, but also
eliminate the grain boundary α phase [16]. However, despite these favorable features, the presence of large
TiB needles (> 50 µm) textured along the prior β grain boundaries in the B-modified Ti64 alloy subjected
to the wire-arc AM process (with relatively lower cooling rate compared to other AM processes) has been
shown to increase the anisotropy in the microstructure and mechanical properties, due to the large TiB
needles being highly susceptible to cracking under tensile loading [21]. Since the size of TiB needles is
dependent on the cooling rate, AM processes with higher cooling rates (e.g., laser powder bed fusion (LPBF)) can lead to the formation of larger numbers of finer TiB precipitates with more homogeneous
distribution, and consequently reduced anisotropy in the microstructure and mechanical properties.
B and C elements are known as the most effective grain refiners in Ti alloys. In addition, their presence in
Ti alloys can lead to the formation of TiB and TiC phases, which can play a promising role in improving
the mechanical properties of Ti alloys [5, 26]. Therefore, minor B4C (as the source of B and C elements)
amount of 0.2wt.% was added to the Ti64 powder in this study to produce the composite powder feedstock.
The composite powder was produced by a novel mechanical mixing strategy, causing the flowability and
apparent packing to be close to those of the monolithic Ti64 powder. A wide range of process parameters
were utilized to fabricate parts out of the composite and monolithic Ti64 powders. The process parameters
leading to the highest densification levels were employed to fabricate compression, and wear test
specimens. Addition of minor B4C to the Ti64 was found to improve the hardness, compressive strengths,
and wear resistance.
2. Materials and Experimental Procedure
2.1. Starting materials and preparation of the composite powder feedstock
The starting powders used in this research were gas atomized Ti64 alloy and B4C with the nominal chemical
compositions reported in [27]. A composite powder feedstock containing 0.2 wt.%B4C (the rest is Ti64) as
the starting reinforcing agent was developed using a novel approach benefitting from the advantages of
both regular mixing and ball milling processes. Production of each 300 g of composite powder by this
technique involved adding 20 g of 2 h-ball milled 3wt.%B4C/Ti64 composite powder to 280 g of monolithic
Ti64 powder, followed by regular mixing for 2 h. The mixing of powders was performed using a highperformance planetary Pulverisette 6 machine operating at a fixed rotational speed of 200 rpm. The ballto-powder weight ratio in the ball milling process was set to be 5:1, while the regular mixing was free from
balls. The metallic balls added to the system in the ball milling process were made of hardened stainless
steel and had a diameter of 10 mm.
2.2. Characterization of the powders
The morphology of the starting powders and the produced composite powders were observed using a Vega
Tescan scanning electron microscopy (SEM) operating at an accelerating voltage of 20 kV. The optical
absorption of Ti64 and B4C powders as well as the developed composite powder was measured using
diffuse reflectance spectroscopy (DRS) technique equipped with an UV-Visible-NIR LAMBDA 950 Perkin
Elmer spectrophotometer. The integrating sphere had a diameter of 150 mm and was coated with Spectralon
with a spectral resolution of 1 nm. A 100% reflectance standard was used as reference to remove the noise.
In order to perform the test, each powder sample was placed in a quartz cuvette and sealed prior to mounting
on a Teflon sample holder. The light sources were Deuterium (D2) and Tungsten with the wavelength
ranges of 200-320 and 320-2500 nm, respectively. Photon Counting photomultiplier tubes (PMT) and Lead
Sulfide (PbS), applicable in the wavelength ranges of 200-860.8 and 861-2500 nm, respectively were used
as detectors. The flowability and packing density of the Ti64 and composite powders were evaluated by the
FT4 Freeman powder rheometer with the procedure thoroughly explained in [27, 28].
2.3. L-PBF processing
An EOS M280 machine (EOS, Krailling, Germany) equipped with a Yb-fiber laser system delivering power
levels of up to 400 W was used in this study. An atmosphere of high purity Ar gas was applied to reduce
the oxygen content in the build chamber and accordingly minimize the oxidation chance. Cubic parts with
the dimensions of 10×10×10 mm3 were printed on a 200 °C-preheated Ti64 build plate (Figure 1(a)). Using
a fixed layer thickness (𝑡) of 30 µm, and a hatch spacing (ℎ) of 100 µm, varying laser powers (𝑃), and
scanning speeds (𝜐) were employed to study the effect of process variables on the quality of the L-PBF
fabricated Ti64 and TMC parts (Table 1). By considering the process parameters, the volumetric energy
density (𝐸𝑣 ) is defined as follows [29]:
𝑃
J
𝐸𝑣 = 𝜐ℎ𝑡 [𝑚𝑚3 ]
Eq. 1
Scanning of layers was conducted using a zigzag scanning strategy, alternating 90° between two successive
layers.
Table 1. The laser powders and scanning speeds employed to fabricate cubic parts.
Laser Power, 𝑷 [W]
Scanning Speed, 𝝊 [mm/s]
𝑬𝒗 [J/mm3]
100, 150, 200, 250
400, 600, 800, 1000
33-208
Figure 1. Schematic view of the fabricated: (a) cubic samples, (b) wear test specimen, and (c) compression test
specimen. As shown in (c), the printed cuboid specimens were subjected to wire electric discharge machining (EDM)
to extract cylindrical compression test samples, as per ASTM-E9-09 standard. (Note: The provided dimensions are
not proportional to the actual size).
The process parameters leading to the highest relative densities in each case (Ti64 and TMC) were
employed to print parts for wear and compression tests (Figure 1(b), and (c)). The discs built for the wear
test were cut off the plate via wire EDM, ground and then polished according to the standard metallography
procedure before testing.
2.4. Microstructural observations
Prior to the microstructural characterization, microhardness measurement and nanoindentation test, the
cubic parts sectioned through the front plane (Figure 1(a)) were ground and polished according to the
standard metallography procedure. The final stage of sample polishing was performed using colloidal silica
with an average particle diameter of 0.06 µm. The non-etched sections were observed using a Keyence
(Osaka, Japan) VHX digital microscope to compare the defects of the parts qualitatively and quantitatively.
The relative density of parts was measured using ImageJ software. The selection of image analysis
technique (instead of Archimedes method) for evaluating the density was due to the fact that the bulk
density needed for relative density measurements of the TMC samples was unknown. The reported relative
density for each sample represents the average of at least 12 measurements. For microstructural studies, the
polished sections were chemically etched using Kroll’s Reagent and were observed using a Nikon optical
microscopy (OM) as well as a Vega Tescan SEM operating at an accelerated voltage of 10 kV. Electron
back-scattered diffraction (EBSD) studies were also carried out to probe the effect of B 4C addition on the
grain size, and texture using hardware and software manufactured by FEI. Spatially resolved EBSD maps
were acquired at the voltage of 20 keV using a step size of 0.3 µm.
2.5.Mechanical testing
The mechanical properties of fabricated parts were evaluated using microhardness and nanohardness
measurements, as well as wear, and room temperature compression tests. Microhardness measurements
were conducted using a Matsuzawa microhardness testing machine with a load of 500 g and a dwell time
of 10 s. The reported microhardness value for each specimen represents the average of at least 5 distinct
measurements. Nanohardness measurements were made using Oliver-Pharr method by utilizing an Anton
Paar NHT3 nano-indentation tester (Anton Paar, Graz, Austria) equipped with a Berkovich pyramidalshaped indenter tip. Before conducting nanoindentation tests, calibration was performed by using a Fused
Silica reference sample. Maximum load of 20 mN, and loading/unloading rate of 40 mN/min were applied
on the polished surface of Ti64 and TMC parts in their front view plane. The pause at the maximum load
was 10 s for all cases. Compression test was conducted at room temperature using an Instron tensile testing
machine as per ASTM E9-19 standard. The strain rate was set to be 0.005 mm/mm/min. The wear
performance of samples was evaluated using an Anton Paar standard TRB3 tribometer in accordance with
the ASTM G99-95a standard [30] at room temperature. The employed ball was made of alumina, and had
a diameter of 6 mm. The applied normal load was 20 N. The linear speed and acquisition rate were set to
be 20 cm/s and 100 Hz, respectively. The tests were performed on three different radii of 4, 6, and 8 mm
on the top surface of each sample. Three distinct spots of the worn track at each radius of each sample were
analyzed with an Alicona microscope to quantify their depth and width and accordingly calculate the wear
rate (𝑊) as follows [31]:
𝑉
𝑛 ×𝑙
𝑊=𝐹
Eq. 2
where 𝑉 is the wear volume, 𝐹𝑛 refers to the applied normal force, and 𝑙 represents the length.
3. Results
3.1. Composite powder feedstock
Figure 2(a) shows the micrograph of the 3wt.%B4C/Ti64 composite powder prepared by the ball milling of
powder constituents for 2 h. As it is evident, Ti64 powder particles in the produced composite powder were
either deformed or almost spherical. The morphological change (spherical to quasi-spherical/flattened) of
the deformed Ti64 particles occurred during the ball milling process, in which the ductile Ti64 powder
particles experienced severe plastic deformation and micro-forging induced by balls. The ball milling
process also provided the attachment of irregular-shape B4C particles to the surface of ductile Ti64 particles,
leading to B4C-decorated Ti64 powder particles. The hard B4C particles hitting the surface of Ti64 ones
during the ball milling stage also caused the roughening of their surface. The 3wt.%B4C/Ti64 composite
powder presented in Figure 2(a) was regularly mixed (balls were absent) with the monolithic Ti64 powder
(Figure 2(b)) to develop the desired composite powder feedstock containing 0.2wt.%B4C. Figure 2(c)
presents the micrograph of this composite powder, in which most of the Ti64 powder particles were fully
spherical and free from B4C particles adhered to their surface since:
(i)
B4C particles were well bonded to the surface of Ti64 particles of the ball-milled
3wt.%B4C/Ti64 composite powder. Accordingly, they did not find the chance to attach to the
Ti64 powder particles added to the system in the regular mixing stage (second stage).
(ii)
The attachment of B4C particles to the Ti64 powder particles requires an external load which
is the impact energy provided by the balls. Due to the absence of balls and consequently the
ball-induced impacts in the regular mixing process, this adherence was not facilitated.
The variation in the optical absorption of the Ti64 and B4C powders as well as the developed
0.2wt.%B4C/Ti64 composite powder is given in Figure 2(d) as a function of the wavelength. At 1070 nm,
which is the wavelength of the L-PBF machine used in this work, Ti64 and B4C powders had optical
absorptions of 66 and 83%, respectively. The composite powder showed a laser absorption of 68%, which
was slighly higher than that of the Ti64 powder. This can be ascribed to the contribution of two factors:
(i)
Addition of only a low content (0.2wt.%) of a highly absorptive powder (B4C) to the Ti64
powder to produce the composite powder. Based on the mixture rule principle, this addition
enhances the absorption of the composite powder since a fraction of the Ti64 powder
constituent is replaced by a higher absorptive material [10, 32].
(ii)
The modest change in the morphology (shape and surface roughness) of Ti64 powder particles
in the portion of composite powder subjected to the ball milling process. This slight deviation
in the morphology acted to increase the frequency of interactions and consequently the optical
absorption [33].
Figure 2. SEM micrographs of: (a) 3wt.%B4C/Ti64, (b) monolithic Ti64, and (c) 0.2wt.%B4C/Ti64 powders. (d)
Optical absorption versus the wavelength for the Ti64 and B4C powders as well as the developed 0.2wt.%B4C/Ti64
composite powder. The wavelength of 1070 nm represents the laser wavelength of the L-PBF machine used in this
study to fabricate parts.
The results of the FT4 Freeman powder rheometer test for Ti64 and 0.2wt.%B4C/Ti64 powders are listed
in Table 2. As can be seen, incorporation of 0.2wt.%B4C powder to the Ti64 powder led to an increase in
the specific energy and adversely affected the flowability (the higher the specific energy, the lower the
flowability). Although B4C particles decorated the surface of only a minor fraction of Ti64 powder particles
in the developed composite powder, they reduced the flowability by increasing the inter-particle friction
and acting as mechanical inter-locking sites during the flow of powder particles [27, 34]. However, it should
be noted that since most of Ti64 particles maintained their desired spherical shape, the produced composite
powder is believed to possess a superior flowability compared to the same composite powder that could be
obtained by a single stage of ball milling for 2 h. The developed composite powder also showed a lower
conditioned bulk density (CBD) than that of the monolithic Ti64 powder. This can be attributed to two
factors: (i) addition of a less-dense material (B4C) to the Ti64 powder, and (ii) the increased inter-particle
friction caused by the presence of irregular-shape B4C particles [27].
Table 2. Flowability and packing density of powders obtained by the FT4 Freeman powder rheometer technique.
Flowability
Density
Specific Energy, SE [mJ/g]
Conditioned Bulk Density, CBD [g/mL]
Ti64
2.1 ± 0.09
2.7 ± 0.02
0.2wt.%B4C/Ti64
3.07 ± 0.05
2.38 ± 0.01
Powder
3.2. Densification level and processability
Figure 3 shows the variation in the density of the L-PBF fabricated Ti64 and TMC parts as a function of
the 𝐸𝑣 . In both cases, the density first increased, and then followed a decreasing trend by increasing the 𝐸𝑣 ,
leading to the maximum densities being achieved within an optimum range of 𝐸𝑣 . The measurements also
revealed that a slightly higher 𝐸𝑣 was required in the TMC case to obtain the maximum density (50 J/mm3
for the TMC vs. 42 J/mm3 for the Ti64 system). To gain a better understanding of the effect of process
parameters on the defect formation, non-etched cross-sections of the front view are provided in Figure 4
and Figure 5 for Ti64 and TMC samples, respectively. The porosities observed in samples with relatively
low 𝐸𝑣 are characterized by their irregular shape (Figure 4(g) and Figure 5(g)), and are known to be caused
by the inter-track and/or inter-layer insufficient overlap. At the laser power of 100 W, the decrease in the
scanning speed (increase in 𝐸𝑣 ) led to lower levels of porosity in both Ti64 and TMC parts (Figure 4(a), (d)
and Figure 5(a), (d)). However, compared to the Ti64 case, the TMC system required a lower scanning
speed (higher 𝐸𝑣 ) to yield an almost defect-free part (600 mm/s vs. 800 mm/s in the case of Ti64). Referring
to Figure 3, the onset of decreasing trend in the density started earlier in Ti64 compared to the TMC scenario
(83 vs. 104 J/mm3). Moreover, a wider range of 𝐸𝑣 led to parts with densities>99% in the TMC system. As
can be descerned in Figure 3, at 𝐸𝑣 levels higher than the optimum range for each case, TMCs possessed
relative densities higher than those of Ti64 counterparts at any given 𝐸𝑣 (see Figure 4(f), and Figure 5(f)).
This suggests that some of the 𝐸𝑣 values which are higher than the optimum value for Ti64, are still within
the optimum processing window for the TMC system. The porosities formed at relatively high 𝐸𝑣 levels
featured a spherical/semi-spherical shape (Figure 4(c), and Figure 5(c)), giving evidence of the keyhole
mode as the defect formation mechanism under this condition [35, 36].
Figure 3. The variation of relative density versus the volumetric energy density (𝑬𝒗 ) for Ti64 and TMC parts. The
designated regions refer to the 𝑬𝒗 range in which highly dense samples were achieved.
Figure 4. Non-etched OM images from the front view cross-sections of Ti64 parts fabricated by different sets of laser
powders and scanning speeds.
Figure 5. Non-etched OM images from the front view cross-sections of TMC parts fabricated by different sets of laser
powders and scanning speeds.
3.3. Microstructure evolution and phase analysis
Figure 6(a), and (f) show 3D view microstructures of the L-PBF processed Ti64 and TMC samples
subjected to the same 𝐸𝑣 of ~83 J/mm3, respectively. Higher magnification micrographs of the selected
regions from the front and top view sections of Ti64 and TMC parts are also presented in Figure 6(b, c),
and Figure 6(g, h), respectively. Referring to Figure 6(a), and (b), the front and top sections revealed the
formation of long (several millimeter) prior β grains extended over multiple layers along the building
direction. The measurements showed two different size scales for the width of these prior β grains.
Accordingly, they were labeled as “wide” and “narrow” prior β grains with a mean width of 86±5.9 µm,
and 9.4±3.1 µm, respectively. The narrow grains were located between the wide ones. Microstructural
observation of the top view showed that the wide prior β grains had a cubic section and were surrounded
by the narrow ones. Figure 6(d) shows the EBSD orientation map, while Figure 6(e) presents the band
contrast image from the front view section of the Ti64 sample. The microstructure was composed of
colonies of α՛ martensite laths with different inclinations mainly at ~±45° relative to the building direction.
Each colony represents several α՛ laths having the same variant. Based on the Burgers relationship, twelve
different α/α՛ variants can be identified for Ti64, accounting for three α/α՛ variants for each of the four
<111>β directions [37, 38]. The XRD analysis results of the Ti64 sample (Figure 7(c)) also featured a fully
martensitic microstructure, similar to that of the starting Ti64 powder (Figure 7(b)). The formation of a
fully martensitic microstructure in the L-PBF processed Ti64 alloy is due to noticeably high cooling rates
of the L-PBF process which well exceeds the critical cooling rate for the martensitic phase transformation
of this alloy (410 °C/s) [39, 40]. Referring to Figure 6(d), the wide and narrow prior β grains were
distinguishable since α՛ laths transformed from each narrow prior β grain featured a different orientation
and color pattern from those of the neighbor wide grains.
Referring to front and side views of the TMC sample provided in Figure 6(f), and (g), typical semi-elliptical
shape tracks were perceptible due to the employed scanning strategy alternating 90° between subsequent
layers. It is worth noting that although the same etchant was used for both Ti64 and TMC samples, the
consolidated tracks were not visible in the Ti64 part. Microstructural characterization of the TMC specimen
also revealed the formation of elongated alternate wide and narrow prior β grains (Figure 6(f), (g), and (h)).
However, compared to the Ti64 sample, narrow prior β grains showed a larger width in the TMC case, and
were located along the center of the tracks (Figure 6(g)). The mean width for the wide and narrow prior β
grains was measured to be 75.5±6.4 µm, and 19.6±4.1 µm, respectively. The sum of the average width of
the wide and narrow prior β grains was almost equal to the employed hatch spacing (100 µm) in both Ti64
and TMC cases. However, the narrow prior β grains surrounding the wide ones were more distinctive and
had a greater width in the TMC sample compared to the Ti64 one. This can be ascribed to the slight
difference between the heat input of the two samples caused by the difference in the optical absorption and
the effective powder layer thickness of the powder systems. Referring to the top views presented in Figure
6, a chessboard pattern of wide square prior β grains surrounded by narrow β grains along their boundaries
was obtained. The same chessboard pattern has been previously observed in several research studies dealing
with the L-PBF processing of the Ti64 alloy. The formation of this pattern has been attributed to the
employed scanning strategy in which the scan directions rotated 90° between subsequent layers [41, 42].
Similar to the Ti64 sample, a fully martensitic matrix was obtained in the L-PBF fabricated TMC system
(Figure 6(i), (j)). However, the measurements gave evidence of finer α՛ laths in the TMC part. The mean α՛
lath width was 1.15±0.15, and 0.92±0.12 µm, while the mean length was 15.6±3.1 and 8.4±7.3 µm for
the Ti64 and TMC samples, respectively. The size of the α/α՛ laths is an important microstructural feature
which governs the mechanical properties of Ti alloys by defining the effective slip length. Smaller α/α՛
colonies generally bring about improved mechanical properties [43, 44]. The XRD phase identification of
the TMC part indicated no peaks corresponding to the B4C phase (Figure 7(a), (d)). Given the fact that
microstructure of L-PBF fabricated TMC was also free from B4C (Figure 6), it can be concluded that these
particles have experienced complete melting/dissolution during the process. The XRD analysis of the TMC
part also revealed a fully martensitic microstructure, the same as that of the Ti64 sample (Figure 7(c), (d)).
In addition, no peaks corresponding to the TiB and TiC phases were detected.
Figure 8 shows SEM micrographs of the TMC part in which fine features are perceptible in the whole
microstructure. Due to their needle-like morphology, these precipitates can be characterized as the TiB
phase formed through the reaction between B and Ti elements. Higher magnification micrographs of the
track interiors (Figure 8(c)) and boundaries (Figure 8(d)) revealed that the fine needle-shape TiB
precipitates were slightly coarser and had a larger spacing in the boundaries than interiors. This
microstructural difference is believed to be the main reason behind observing the tracks in the TMC part,
as opposed to the Ti64 counterpart. Within the track interior, TiB needles were mainly oriented toward the
center of the melt pool (Figure 8(c)). This directionality is known to be due the unidirectional heat flux
occurring from center toward the boundaries of the melt pool [59, 73, 74], allowing preferential growth of
these needle-shape features opposite to the heat flux direction. It is noteworthy that the absence of TiB
phase in the XRD pattern of the TMC part (Figure 7) might be due to its noticeably small size.
Figure 6. (a) OM 3D view of the microstructure of the L-PBF fabricated Ti64; (b), (c) higher magnification OM
images of the selected areas of the front and top views shown in (a), respectively; (d) inverse pole figure (IPF-Z), and
(e) band contrast EBSD maps of the front view of the Ti64 sample; (f) OM 3D view of the microstructure of the LPBF fabricated TMC part; (g), (h) higher magnification OM images of the selected areas of the front and top views
shown in (f), respectively; (i) inverse pole figure (IPF-Z), and (j) band contrast EBSD maps of the front view of the
TMC sample. The employed laser powder, and scanning speed for both Ti64 and TMC samples were 250 W, and
1000 mm/s, respectively.
Figure 7. XRD patterns of the starting: (a) B4C, and (b) Ti64 powders, as well as the L-PBF fabricated: (c) Ti64, and
(d) TMC parts. Both Ti64 and TMC samples were manufactured using a laser power of 250 W and a scanning speed
of 1000 mm/s.
Figure 8. Secondary electron (SE)-SEM images from the front view of the L-PBF fabricated TMC part. (a) General
view in which the cross-section of a track has been highlighted. Higher magnification micrograph of the selected
dashed square in (a) is shown in (b). The square regions marked as “A” and “B” in (b) are shown in a higher
magnification in (c) and (d), respectively.
3.4. Mechanical Properties
To gain insights into the level of improvement in the mechanical properties caused by the composite
fabrication, the hardness, strength, ductility, and wear performance of L-PBF fabricated TMC was
evaluated and compared to those of the Ti64 counterpart. As shown in Figure 9, the developed TMC showed
5% higher microhardness than that of the Ti64 part. However, nanoindentation measurements revealed a
greater difference between the nanohardness values (25%). It is worth noting that compared to the Ti64
part, a smaller standard deviation was detectable in the nanohardness measurements of the TMC sample.
This can be attributed to the enhanced microstructure homogeneity caused by the uniform distribution of
fine TiB precipitates in the matrix of the TMC sample.
The typical compressive stress-strain curves of the Ti64 and TMC parts are presented in Figure 9(b). Table
3 summarizes the most important properties of these two materials obtained from several tests performed
on each material. As it is evident, both samples had almost the same ultimate strength
(𝜎𝑢 ) and fracture strain (𝑒𝑓 ). However, the TMC sample possessed ~8% higher yield strength (𝜎𝑦 ) than the
Ti64 counterpart. Therefore, although the incorporation of nano-scale/sub-micron TiB reinforcement into
the Ti64 matrix brought about higher hardness, and strength, it did not have an adverse effect on the
ductility.
Figure 9. (a) The microhardness and nanohardness values, and (b) typical compressive engineering stress-strain curves
of the L-PBF fabricated Ti64 and TMC parts.
Table 3. Mechanical properties of the L-PBF fabricated Ti64 and TMC parts obtained by compression testing.
Sample
𝜎𝑦
𝜎𝑢
Young’s Modulus, E [GPa]
Fracture Strain, 𝑒𝑓 [%]
Ti64
TMC
1190
1284
1833
1850
114.4
119.3
21
20.6
Figure 10(a), and (b) shows the variation in the coefficient of friction (COF) for the L-PBF fabricated
optimum Ti64 and TMC parts along the sliding distance, respectively. The listed in Table 4, there was no
significant difference between the COF for the Ti64 and TMC samples. However, the analysis of the wear
tested samples revealed a noticable difference between the depth and width of the worn tracks left on the
surface of Ti64 and TMC parts. Figure 10(c), and (d) illustrates representative depth profiles along the
width of the worn tracks for Ti64 and TMC samples across the directions shown in Figure 10(d), and (e),
respectively. As observed, compared to the Ti64 sample, the TMC featured a shallower depth and a
narrower width. The wear rate of samples was calculated by measuring the dimensions of the worn tracks
at several spots of each radii and consequently calculating the wear volume. Referring to in Table 4, the
TMC sample showed 15% lower wear rate than the Ti64 one. This reveals the remarkable influence of
incorporated TiB reinforcements on the improved wear resistance of the L-PBF fabricated TMC.
Figure 10. (a), (b) the variation of COF versus sliding distance for Ti64 and TMC samples, respectively.
Representative depth-width profile of the worn tracks for (c) Ti64, and (d) TMC samples obtained along the paths
depicted in the 3D surface topography images in (e), and (f), respectively.
Table 4. The coefficient of friction (COF) and wear rate of L-PBF processed Ti64 and TMC samples.
System
COF
Wear rate [mm3/(N.m)]
Ti64
0.4218
1.318×10-3
TMC
0.4314
1.115×10-3
Figure 11 shows the SEM micrographs of the wear tracks for Ti64 and TMC samples. The wear tracks of
both samples presented wear scars along the sliding direction which is typical of the wear surfaces. In
addition, ploughing grooves were observed in both cases. As can be seen in the higher magnification
micrographs in Figure 11(b), and (d), the wear surface contained delamination cracks in the Ti64 and TMC
samples. These cracks are generated as a cause of the high strain levels experienced by the surface of the
sample when being in contact with the pin during sliding. As indicated in Figure 11(b), and (d), such high
strain levels can bring about the delamination of the material during the wear test. A close examination of
the wear tracks at several spots of each radii (4, 5, and 8 mm) revealed that compared to the Ti64 sample,
the delamination cracks and delamination layers were less intense in the TMC sample. This can be due to
the higher hardness and plastic deformation resistivity of the TMC sample caused by the strengthening
effects of TiB reinforcements. It has been shown that owing to its high modulus and hardness, incorporation
of slight amount of TiB into the Ti alloy matrix can lead to a significant increase in the wear resistance [26].
Figure 11. SEM images of the wear tracks after the sliding wear test for the L-PBF fabricated: (a), (b) Ti64; and (c),
(d) TMC samples. Higher magnification micrographs of the selected regions in (a), and (c) and presented in (b), and
(d), respectively.
4. Discussions
4.1. Processability
As the density measurement results and microstructural observations revealed, addition of a minor amount
of B4C (0.2wt.%) to the Ti64 powder changed its L-PBF processability (Figure 3, Figure 4, and Figure 5).
This clearly reveals that not only the process parameters but also the powder characteristics and laserpowder interactions need to be re-examined when dealing with a composite powder system containing even
minor amounts of a second constituent. As proposed in a very recent study on L-PBF processing of VC/H13
system [45], the volumetric energy density calculation equation most frequently used in L-PBF process
needs to be revisited for composite powder systems by taking the laser absorptivity (𝐴) and relative powder
bed packing density (𝜌𝑏𝑒𝑑 ) of the powder also into account as follows:
𝑒𝑣 = (𝐴 × 𝜌𝑏𝑒𝑑 )𝐸𝑣
Eq. 3
where 𝑒𝑣 represents the modified volumetric energy density, and 𝜌𝑏𝑒𝑑 is defined as 𝐶𝐵𝐷/𝜌𝑏𝑢𝑙𝑘 (𝜌𝑏𝑢𝑙𝑘
signifies the bulk density of the material). This equation suggests that powders with higher (𝐴 × 𝜌𝑏𝑒𝑑 )
require lower 𝐸𝑣 to yield the optimum 𝑒𝑣 . Referring to the CBD measurements results provided in Table 2,
optical absorption values presented in Figure 2(a), and assuming 𝜌𝑏𝑢𝑙𝑘 to be 4.43 g/cm3 for both Ti64 and
TMC bulk samples, the (𝐴 × 𝜌𝑏𝑒𝑑 ) term for the Ti64 and 0.2wt.%B4C/Ti64 systems is 0.4 and 0.36,
respectively. This justifies the need for slightly higher 𝐸𝑣 in the composite system (compared to the Ti64)
to reach the highest densification level.
The observed increasing and then decreasing variation in the density of the Ti64 and TMC samples (Figure
3) is in agreement with the results reported in the literature for the L-PBF processing of metallic materials
[46, 47]. The decline in the relative density at relatively high 𝐸𝑣 values exceeding the optimum value has
been well documented in the literature and has been shown to be due to the transition of melting mode from
conduction to keyhole [29, 35, 48, 49]. During the keyhole mode of melting, the material evaporation
caused by application of high 𝐸𝑣 leads to the development of a vapor cavity which further enhances the
laser absorption and enables the laser beam to penetrate to a deeper depth. The collapse of the cavity can
leave voids in the material which are known as the keyhole porosities and are spherical in shape [48, 50].
As it is perceptible in Figure 3, the decreasing trend in the relative density occurred at a slightly lower 𝐸𝑣
in the Ti64 system compared to the TMC case (83 vs. 104 J/mm3). One of the most effective estimations
for the criteria of keyhole threshold has been shown to be based on the normalized enthalpy (
𝛥𝐻
)
ℎ𝑠
as follows
[48]:
𝛥𝐻
ℎ𝑠
=
𝐴𝑃√𝐷
𝑘𝑇𝑚 √𝜋𝑣𝑑 3
≥
𝜋𝑇𝑏
𝑇𝑚
Eq. 4
in which 𝛥𝐻 is the specific enthalpy, ℎ𝑠 is the enthalpy at melting defined as
𝑘𝑇𝑚
, 𝑇𝑚
𝐷
is the melting point,
𝑇𝑏 is the boiling point, 𝑑 is the laser beam size, 𝑘 is the thermal conductivity, and 𝐷 is the thermal
diffusivity. Eq. 4 takes into account the effect of the applied 𝑃, 𝑣, and 𝑑 on the threshold for the formation
of keyhole melting mode. However, it fails to consider the influence of layer thickness as one of the
important parameters playing role on the melting mode. Eq. 5 provides the modified threshold for keyhole
mode of melting by taking the effective layer thickness also into account [51, 52].
𝛥𝐻
ℎ𝑠
0.865𝐴𝑃𝐷
𝑚 √𝜋 𝑑𝑡𝑒𝑓𝑓 𝑣
= 𝑘𝑇
≥
𝜋𝑇𝑏
𝑇𝑚
Eq. 5
Eq. 5 can be also rewritten as:
𝑃
𝑣
≥
𝑘𝑑𝑇𝑏 𝑡𝑒𝑓𝑓 √𝜋3
Eq. 6
0.865𝐴𝐷
Assuming the same 𝑘, and 𝑇𝑏 for the Ti64 and TMC systems, and considering 𝑡𝑒𝑓𝑓 = 𝜌
𝑡
𝑏𝑒𝑑
𝑃
𝑣
1
∝ 𝐴×𝜌
Eq. 7
𝑏𝑒𝑑
Referring to Eq. 7, the critical
[34]:
𝑃
𝑣
for the keyhole threshold in the composite system is expected to be 10%
higher than that of the Ti64 case. However, based on the experimental results, this difference was ~25%.
Although a portion of this discrepancy may have raised because of assuming the same 𝑘, and 𝑇𝑏 for both
systems, there are still other factors related to the composite systems which have not been considered in the
currently existing keyhole threshold models. For instance, given the fact that the dissolution of B4C particles
during the L-PBF processing of the composite system is an endothermic reaction, a part of the energy
absorbed by the powder bed is consumed by their dissolution.
4.2. Microstructure evolution
As shown in Figure 6, the microstructure of both Ti64 and TMC parts contained elongated columnar prior
β grains which were extended along the building direction. The formation of such columnar prior β grains
is a major characteristic of the L-PBF processed Ti64 alloy parts caused by the combination of epitaxial
and competitive growth during the solidification. The epitaxial growth deals with the solidification of the
melt pool through the growth (rather than nucleation) of β crystals on the β grains of the adjacent
track/underlying layer. Those prior β grains having their easy-growth crystallographic direction (<001>
direction in cubic crystal structures) perpendicular to the fusion line (along the heat flux direction)
experience a preferential growth, meaning that faster growth of these grains crowds out the β grains oriented
in less favorable directions [53-55]. As a result, the final solidification structure consists of columnar prior
β grains extended along the building direction.
As the laser beam interacts with the 0.2wt.B4C/Ti64 composite powder bed, Ti64 powder particles
experience melting since their melting temperature is well below the temperature levels induced during the
L-PBF process. However, owing to their noticeably high melting point of 3036 K [56] as well as the
extremely fast cooling rates associated with the L-PBF process (104-106 K/s), complete melting of B4C
particles is rather difficult or even impossible. As the microstructural characterizations (Figure 6 and Figure
8) and XRD analysis results (Figure 7) revealed, the fabricated TMC did not contain B4C constituent in the
final microstructure. In fact, B4C particles fully reacted with the surrounding melt. This reaction is believed
to occur through the dissolution of B4C particles rather than their melting. The full dissolution of B4C
particles in the surrounding molten material is an interesting phenomenon which may not be expected to
take place due to the extremely high cooling rate of the L-PBF process and consequently the noticeably
short time available for diffusion-based dissolution mechanism. The fast and complete dissolution of highmelting-point B4C particles in this study can be attributed to: (i) their relatively low fraction and small size,
and (ii) convective flows (e.g., Marangoni flow, etc.) active within the melt pool during the L-PBF process
which consecutively provide fresh molten material (not enriched from B atoms) adjacent to the solid B4C
particles [40]. Based on the calculations, the molten material formed by complete dissolution of B4C
constituent during the L-PBF processing of the 0.2wt.%B4C/Ti64 composite powder contains 0.156 and
0.039 wt.% B and C, respectively.
Figure 12(a) shows the equilibrium phase diagram of the Ti64-B system where the isopleth for the Ti640.156wt.%B system is specified. The schematic illustration of the solidification sequence of this isopleth is
also depicted in Figure 12(e)-(h), in which the solidification starts with the formation of β phase from the
liquid and their subsequent growth (Figure 12(e), and (f)). Due to the extremely low/negligible solubility
of B and C in the β phase, they are rejected from the growing solidification interface upon cooling between
the liquidus and eutectic temperatures (Figure 12(a), (f), and (g)), leading to the progressive enrichment of
the liquid from B and C. Since the C in the investigated TMC system is noticeably lower than the
equilibrium solubility limit of this element in Ti (0.039 vs. 0.125 wt.%), the C content of the enriched liquid
is still below this limit, causing C atoms to exist as solid solution in the Ti64 matrix. At the eutectic
temperature, the remaining liquid enriched from B transforms to eutectic 𝑇𝑖𝐵+ 𝛽 (𝑇𝑖𝐵𝐸 +𝛽𝐸 ) through a
pseudo-eutectic reaction (Figure 12 (a), (g), and (h)). Therefore, the equilibrium solidification sequence can
be summarized as:
𝐿 → 𝛽 + 𝐿 → 𝛽 + (𝑇𝑖𝐵𝐸 + 𝛽𝐸 )
The β phase transforms to (α+β) at the β transus temperature. Therefore, equilibrium microstructure at room
temperature consists of lamellar (α+β) islands surrounded by narrow band of (𝑇𝑖𝐵𝐸 + 𝛽𝐸 ).
The isopleth of the Ti64 alloy is presented in Figure 12(b), where the solidification starts with the nucleation
of β phase from the liquid followed by their growth. Despite the Ti64-0.156wt.%B system with a relatively
wide solidification range in the equilibrium state, that of the Ti64 alloy is noticeably narrow. This suggests
that no significant partitioning is occurred in the liquid during the solidification of the Ti64 alloy.
Figure 12. (a) Equilibrium phase analysis diagram of the Ti64-B system with the B content in the range of 0-0.3 wt.%.
The provided isopleth represents the investigated 0.2wt.%B4C/Ti64 system which contains 0.156wt.%B. (b)
Equilibrium phase analysis diagram of the Ti-6Al-XV, in which the isopleth of the Ti-6Al-4V (Ti64) alloy has been
shown. Scheil-Gulliver solidifications models of the: (c) Ti64-0.2wt.%B4C, and (d) Ti64 systems. The inset in (c)
shows the variation of the TiB fraction within the specified temperature range (1530-1420 °C). (e)-(h) Schematic
illustration of the solidification sequence of the 0.2wt.%B4C/Ti64 system in the equilibrium condition. The liquid (L)
becomes enriched from B and C as the solidification progresses.
As the microstructural observations revealed, the fabricated TMC contained fine nano-scale/sub-micron
TiB needles homogeneously dispersed in the martensitic matrix. This microstructure differs from the
microstructure predicted by the equilibrium phase diagram in Figure 12(a), where the large prior β grains
are surrounded by the pseudo-eutectic reaction products (𝑇𝑖𝐵𝐸 +𝛽𝐸 ) along their boundaries (Figure 12(h)).
Although the formation of a martensitic matrix (rather than the (α+β) two-phase microstructure) in the LPBF fabricated TMC can be justified based on the extremely fast cooling rates of the process, the reason
behind the deviation of the solidification path from the equilibrium condition needs further analysis.
Figure 12(c) illustrates the solidification sequence proposed by the non-equilibrium Scheil-Gulliver model
for the TMC system investigated in this study. This model accounts for the infinitely fast diffusion in the
liquid and no diffusion in the solid. As can be seen, the Scheil-Gulliver model predicts the same
solidification sequence as that of the equilibrium condition, revealing that even the non-equilibrium ScheilGulliver solidification model fails to predict the solidification sequence of the TMC system subjected to
the L-PBF process. However, unlike the equilibrium condition in which the TiB phase is formed at a
specific temperature (Eutectic temperature of 1440 °C), the Scheil-Gulliver model predicts the formation
of TiB to occur within a relatively wide range starting at 1530 °C and finishing at 1440 °C. This can be
ascribed to the lack of solute diffusion in the solid based on the Scheil-Gulliver model which provides the
liquid with the B content required for the formation of TiB at higher temperatures.
Figure 13(a), and (b) shows the microstructure of the last solidified layer of the fabricated TMC part, which
is composed of the TiB needles homogeneously dispersed in a martensitic matrix. The microstructure
evolution of the last layer is schematically illustrated in Figure 13(c)-(g). Although B has a negligible
solubility in the equilibrium condition in the high-temperature β phase (a maximum of ~0.05 wt.% (Figure
13(a))), the extremely fast cooling rate associated with the L-PBF process can significantly diminish the
rejection of B atoms into the liquid during the solidification (Figure 13(d), and (e)). For instance, rapid
solidification of the Ti-B systems has been shown to cause supersaturation of B in the matrix up to 10 at.%
(2.5 wt.%B) [57], which is noticeably higher than the B content of the TMC system in this study (0.156
wt.%). Accordingly, the high-temperature prior β phase is supersaturated from B. Due to the decreased
solubility limit during the cooling stage, a portion of the supersaturated B tends to precipitate out as the TiB
phase from both the β and the later transformed B-supersaturated martensite (M) phases (Figure 13(f), and
(g)). Compared to the bulk of the TMC sample shown in Figure 8, lower volume fraction of TiB phase was
perceptible in the last layer. This clearly reveals the significant role of the subsequent thermal cycles (caused
by the heat conduction from the melt pool to the previously deposited tracks or layers) on the precipitation
of TiB from the B-supersaturated matrix. The same phenomenon has been also observed in the powder
metallurgy-processed 1.7wt.%B/Ti64 alloy, where the exposure of the material to elevated temperatures
during the thermo-mechanical processing led to the increase in the volume fraction of the TiB phase in the
microstructure [58].
Both the equilibrium and Scheil-Gulliver non-equilibrium solidification models are based on the scenario
where the temperature gradient is low enough to allow the formation of a constitutionally undercooled zone
in front of the solidification interface, favoring the nucleation of new prior β grains. However, when an
extremely steep temperature gradient exists (e.g., such as that of the L-PBF process), very limited or even
no nucleation events would occur ahead of the solidification interface. This explains the observation that
the TMC part also featured columnar prior β grains with almost the same size as those of the Ti64 alloy
(Figure 6). It is worth noting that the inability of the minor B amounts in avoiding the formation of columnar
prior β grains has been also reported in B-modified Ti64 alloys subjected to the AM processes with thermal
gradients noticeably lower than that of the L-PBF process (WAAM) [59].
Figure 13. (a), and (b) SEM micrographs of the last layer of the L-PBF fabricated TMC part. Higher magnification
micrograph of the selected region in (a) is shown in (b). (c)-(g) Schematic illustration of the microstructure evolution
of the last layer of the L-PBF fabricated TMC part during cooling. “M” in (g) refers to martensite.
As observed in Figure 8 and Figure 13, TiB reinforcements had a needle-like shape and were uniformly
dispersed in the martensitic matrix. The reason behind such a morphology has been traced back to the strong
preferential growth along the [010]27 direction compared to the other directions [60]. The average length
and width of TiB needles were 1.6 µm and 260 nm, respectively, being significantly smaller than those
formed in Ti-B systems with trace B content subjected to the DED [61], direct laser deposition (DLD) [62],
and wire-arc AM (WAAM) [21] processes, which usually have a length in the order of several microns.
This corresponds to the noticeably higher cooling rates associated with the L-PBF process. The relatively
large TiB needles have been reported to be detrimental to the mechanical properties since they can act as
crack nucleation sites during loading [16, 21, 61]. Therefore, the L-PBF fabricated TiB reinforced TMCs
are more favorable since they benefit from greater number of finer TiB reinforcements with smaller interreinforcement spacing uniformly dispersed in the matrix (as opposed to fewer numbers of TiB
reinforcements with a larger spacing).
4.3. Mechanical Properties
Based on the hardness measurements, compression test results, and wear test, the TMC sample benefitted
from a higher hardness, yield strength, and wear resistance than the Ti64 counterpart. The improvement in
the mechanical properties of the TMC over the Ti64 can be ascribed to the contribution of several
strengthening mechanisms, including the grain refinement, second constituent, dislocation, and solid
solution. In the following, the level of enhancement in the yield strength caused by each mechanism is
discussed.
Grain refinement: The increased fraction of boundaries obtained by grain refinement elevates the plastic
deformation resistivity and consequently improves the strength of the material [63, 64]. Based on the HallPetch relationship, the increase in the yield strength of the TMC caused by the grain refinement can be
expressed as follows [63]:
∆𝜎𝐻𝑃 = 𝑘𝐻𝑃 (
1
√𝑑2
−
1
√𝑑1
)
Eq. 8
in which 𝑘𝐻𝑃 is the Hall-Petch coefficient for Ti64; while 𝑑2 , and 𝑑1 are the mean grain sizes of the TMC
and Ti64 samples, respectively. Eq. 8 is typically used for the equiaxed microstructures whereas the Ti64
and TMC samples featured martensitic microstructures characterized by α՛ laths. Therefore, the average
width of α՛ laths in each material was taken into account [65].
Second constituent strengthening: The strength improvement caused by the presence of TiB reinforcements
in the TMC part is defined as [66]:
1
𝑙
∆𝜎𝑇𝑖𝐵 = 2 𝜎𝑦𝑚 × 𝑉𝑇𝑖𝐵 × 𝑑𝑇𝑖𝐵 𝜔0
Eq. 9
𝑇𝑖𝐵
where 𝜎𝑦𝑚 is the yield strength of the Ti64 matrix, 𝑉𝑇𝑖𝐵 is the volume fraction of TiB, (
𝑙𝑇𝑖𝐵
)
𝑑𝑇𝑖𝐵
refers to the
aspect ratio of TiB (treated as cylinders with the average length and width of 𝑙 𝑇𝑖𝐵 , and 𝑑𝑇𝑖𝐵 , respectively),
and 𝜔0 is the orientation factor which has a value between 0 and 1. Due to the random distribution of TiB
whiskers in the matrix of the TMC part, 𝜔0 is equal to 0.125 or 0.27 based on the 3D or 2D random array
model, respectively [67].
Dislocation strengthening: The increment in the strength originated from the dislocation strengthening can
be estimated as follows [68]:
2
∆𝜎𝑑𝑖𝑠 = √(∆𝜎𝑜𝑟𝑜 )2 + (∆𝜎𝑡ℎ𝑒 )2 + (∆𝜎𝑔𝑒𝑜 )
Eq. 10
in which ∆𝜎𝑜𝑟𝑜 is the strength improvement by the Orowan mechanism, ∆𝜎𝑡ℎ𝑒 represents the stress
enhancement caused by the thermal expansion mismatch between the TiB and the Ti64 matrix, and ∆𝜎𝑔𝑒𝑜
signifies the stress increment due to the geometrically necessary dislocations (GNDs).
In the composites containing reinforcements not exceeding 1 µm in size (e.g. TiB needles with
nanoscale/submicron width in this study), the reinforcements can act as strong obstacles to the movement
of dislocations and impede their motion on the matrix’s slip plane [69]. The strengthening effect caused by
this phenomenon is known as the Orowan mechanism, in which the shear component of the applied stress
bows out the dislocation between reinforcements. This produces a back stress which in turn prevents the
migration of further dislocations and consequently enhances the yield strength [70]. The Orowan stress can
be described by the Orowan-Ashby equation as [71]:
∆𝜎𝑜𝑟𝑜 =
0.13 𝐺𝑚 𝐵𝑚
𝐷
ln 2𝑏𝑇𝑖𝐵
𝜆
𝑚
Eq. 11
where 𝐺𝑚 is the shear modulus of the Ti64 matrix, 𝐵𝑚 is the Burgers vector of the Ti64 matrix, 𝐷𝑇𝑖𝐵 is the
equivalent diameter of the TiB reinforcement, and 𝜆 is the interparticle spacing. 𝐷𝑇𝑖𝐵 and 𝜆 can be
calculated as [68, 72]:
3
𝐷𝑇𝑖𝐵 = √1.5𝑑2𝑇𝑖𝐵 𝑙 𝑇𝑖𝐵
Eq. 12
1
𝜆 ≈ 𝐷𝑇𝑖𝐵 [(2𝑉𝑇𝑖𝐵 )−3 − 1]
Eq. 13
Since the thermal expansion coefficients of TiB and Ti64 are very close to each other, ∆𝜎𝑡ℎ𝑒 can be
neglected.
∆𝜎𝑔𝑒𝑜 deals with the strength improvement caused by the GNDs required to accommodate the plastic
deformation mismatch between the TiB reinforcement and the matrix, and is expressed as follows [73]:
𝜀𝑇−𝑚 𝑏𝑚 𝑉𝑇𝑖𝐵
𝐷𝑇𝑖𝐵
∆𝜎𝑔𝑒𝑜 = 0.4𝐺𝑚 √
Eq. 14
in which 𝜀𝑇−𝑚 is the compressive true strain of the matrix at the yield point.
Solid solution strengthening: As the microstructural observations and phase analysis results revealed, the
TiC phase was not detected in the fabricated TMC part. Since the C content in the studied TMC is less than
the solid solubility limit of C in Ti, C atoms have been entrapped in the matrix as solid solution. Given the
fact that C is an interstitial alloying element, it can generate a significantly strong obstacle (compared to
the substitutional elements) for the movement of dislocations in the matrix and consequently lead to higher
levels of solid solution strengthening [74]. The solid solution strengthening caused by the interstitial solutes
is generally proportional to the square root of the concentration as [75]:
∆𝜎𝑠𝑠 = 0.002𝐺𝑚 𝑐1/2
Eq. 15
where 𝑐 is the carbon concentration in at.%.
By taking all the strengthening mechanisms into account, Ramakrishnan’s approach (Eq. 16) can be used
to calculate the yield strength of the TMC component (𝜎𝑦−𝑇𝑀𝐶 ) [76]:
𝜎𝑦−𝑇𝑀𝐶 = 𝜎𝑦−𝑚 (1 +
∆𝜎𝐻𝑃
∆𝜎
∆𝜎𝑑𝑖𝑠
) (1 + 𝑇𝑖𝐵 ) (1 +
) (1
𝜎𝑦−𝑇𝑀𝐶
𝜎𝑦−𝑇𝑀𝐶
𝜎𝑦−𝑇𝑀𝐶
+
∆𝜎𝑠𝑠
)
𝜎𝑦−𝑇𝑀𝐶
Eq. 16
Eq. 16 estimates the yield strength of the TMC sample to be ~1307 MPa, which is in a good agreement
with that obtained by the experiment (1284 MPa). Table 6 summarizes the contribution of each
strengthening mechanism in the yield strength increment of the TMC sample over the Ti64 counterpart.
While the ∆𝜎𝑇𝑖𝐵 is the direct strengthening mechanism originating from the incorporated TiB
reinforcement, ∆𝜎𝐻𝑃 , and ∆𝜎𝑑𝑖𝑠 are the indirect strengthening mechanisms activated due to the presence of
this phase in the microstructure. Therefore, taking both direct and indirect strengthening mechanisms into
consideration, ~70% of the total yield strength enhancement in the TMC is caused by the presence of TiB
phase. On the other hand, the remaining ~30% is linked to the solid solution strengthening of the matrix
induced by C atoms. The improved hardness and wear resistance of the TMC parts can be also attributed
to the same strengthening mechanisms discussed above for the strength.
Table 5. The properties and parameters used in obtaining the contribution of different strengthening mechanisms.
Properties and Parameters
𝑘𝐻𝑃
𝑑2
𝑑1
𝑏𝑚
𝐺𝑚
𝜔0
𝑉𝑇𝑖𝐵
𝑙 𝑇𝑖𝐵
𝑑 𝑇𝑖𝐵
𝜀𝑇−𝑚
𝐷𝑇𝑖𝐵
Value
300 MPa.µm1/2
0.92 µm
1.15 µm
0.295 nm
45 GPa
0.125
0.05
1590 nm
264 nm
0.0122
0.55 µm
Source
[77]
Measured
Measured
[78]
[79]
[67]
Measured
Measured
Measured
Measured
Measured
Table 6. The calculated values of strengthening mechanisms in the TMC part along with their contribution in the
enhancement of the yield strength over the Ti64 counterpart.
Strengthening Mechanism
Value [MPa]
Proportion in the total ∆𝜎 [%]
∆𝜎𝐻𝑃
33
29.1
∆𝜎𝑇𝑖𝐵
22.4
19.8
∆𝜎𝑑𝑖𝑠
21.3
18.8
∆𝜎𝑠𝑠
36.6
32.3
5. Conclusions
In this study, the laser powder bed fusion (L-PBF) process was implemented to fabricate high performance
titanium matrix composite (TMC) parts using a composite powder feedstock produced by a novel technique.
The main conclusions of this research study can be summarized as follows:
1- The proposed composite powder preparation technique benefiting from the advantages of both
regular mixing and ball milling was found to yield a 0.2wt.%B4C/Ti64 composite powder having
the flowability and apparent packing density close to those of the monolithic Ti64 powder.
2- In both Ti64 and TMC cases, an increasing and then decreasing trend in the relative density of parts
was achieved by enhancing the volumetric energy density. Compared to the Ti64 case, slightly
higher energy densities were needed in the TMC to achieve almost fully dense samples. Moreover,
the decline in the relative density took place at higher energy densities in TMC than Ti64.
3- Microstructural characterizations revealed the formation of large columnar prior β grains in both
Ti64 and TMC samples which were elongated along the building direction. The TMC sample
featured nano-scale/sub-micron TiB needles homogeneously dispersed in a fully martensitic
matrix.
4- The equilibrium and Scheil-Gulliver non-equilibrium solidification models were found not to be
able to predict the microstructure evolution experienced by the L-PBF fabricated TMC parts. This
was attributed to the extremely high cooling rates of the L-PBF process which led to the
supersaturation of B atoms in the β grains during solidification.
5- Although showing almost the same ultimate strength and fracture strain as those of the Ti64
counterpart, the TMC part possessed 5% higher microhardness, 8% higher yield strength, and 12%
lower wear rate. The improved yield strength of the TMC was justified by calculating the
contribution of different strengthening mechanisms involved. While ~70% of this improvement
was due to the direct and indirect strengthening effects caused by the TiB phase, the remaining
~30% was originating from the solid solution strengthening of the matrix by C atoms.
References
[1] D. Miracle, Metal matrix composites–from science to technological significance, Composites science and
technology 65(15-16) (2005) 2526-2540.
[2] S.C. Tjong, Recent progress in the development and properties of novel metal matrix nanocomposites reinforced
with carbon nanotubes and graphene nanosheets, Materials Science and Engineering: R: Reports 74(10) (2013) 281350.
[3] C. Cui, B. Hu, L. Zhao, S. Liu, Titanium alloy production technology, market prospects and industry
development, Materials & Design 32(3) (2011) 1684-1691.
[4] M.J. Donachie, Titanium: a technical guide, ASM international2000.
[5] Y. Jiao, L. Huang, L. Geng, Progress on discontinuously reinforced titanium matrix composites, Journal of
Alloys and Compounds 767 (2018) 1196-1215.
[6] M.G. Glavicic, P.A. Kobryn, F. Spadafora, S.L. Semiatin, Texture evolution in vacuum arc remelted ingots of
Ti-6Al-4V, Materials Science and Engineering: A 346(1) (2003) 8-18.
[7] G. Singh, U. Ramamurty, Boron modified titanium alloys, Progress in Materials Science 111 (2020) 100653.
[8] S. Tamirisakandala, R.B. Bhat, J.S. Tiley, D.B. Miracle, Grain refinement of cast titanium alloys via trace boron
addition, Scripta materialia 53(12) (2005) 1421-1426.
[9] E. Fereiduni, M. Yakout, M. Elbestawi, Laser-based additive manufacturing of lightweight metal matrix
composites, Additive Manufacturing of Emerging Materials, Springer2019, pp. 55-109.
[10] E. Fereiduni, M. Elbestawi, Process-structure-property relationships in additively manufactured metal matrix
composites, Additive Manufacturing of Emerging Materials, Springer2019, pp. 111-177.
[11] R.R. Sélo, S. Catchpole-Smith, I. Maskery, I. Ashcroft, C. Tuck, On the thermal conductivity of AlSi10Mg and
lattice structures made by laser powder bed fusion, Additive Manufacturing 34 (2020) 101214.
[12] A. Salmi, F. Calignano, M. Galati, E. Atzeni, An integrated design methodology for components produced by
laser powder bed fusion (L-PBF) process, Virtual and Physical Prototyping 13(3) (2018) 191-202.
[13] E. Fereiduni, A. Ghasemi, M. Elbestawi, Selective Laser Melting of Aluminum and Titanium Matrix
Composites: Recent Progress and Potential Applications in the Aerospace Industry, Aerospace 7(6) (2020) 77.
[14] M. Narvan, A. Ghasemi, E. Fereiduni, S. Kendrish, M. Elbestawi, Part deflection and residual stresses in laser
powder bed fusion of H13 tool steel, Materials & Design 204 (2021) 109659.
[15] S.M.J. Razavi, F. Berto, Directed energy deposition versus wrought Ti‐ 6Al‐ 4V: a comparison of
microstructure, fatigue behavior, and notch sensitivity, Advanced Engineering Materials 21(8) (2019) 1900220.
[16] K. Zhang, X. Tian, M. Bermingham, J. Rao, Q. Jia, Y. Zhu, X. Wu, S. Cao, A. Huang, Effects of boron addition
on microstructures and mechanical properties of Ti-6Al-4V manufactured by direct laser deposition, Materials &
Design 184 (2019) 108191.
[17] C. Liu, Y. Lu, X. Tian, D. Liu, Influence of continuous grain boundary α on ductility of laser melting deposited
titanium alloys, Materials Science and Engineering: A 661 (2016) 145-151.
[18] B.E. Carroll, T.A. Palmer, A.M. Beese, Anisotropic tensile behavior of Ti–6Al–4V components fabricated with
directed energy deposition additive manufacturing, Acta Materialia 87 (2015) 309-320.
[19] W.J. Sames, K.A. Unocic, R.R. Dehoff, T. Lolla, S.S. Babu, Thermal effects on microstructural heterogeneity
of Inconel 718 materials fabricated by electron beam melting, Journal of materials research 29(17) (2014) 19201930.
[20] Z. Mahbooba, H. West, O. Harrysson, A. Wojcieszynski, R. Dehoff, P. Nandwana, T. Horn, Effect of
hypoeutectic boron additions on the grain size and mechanical properties of Ti-6Al-4V manufactured with powder
bed electron beam additive manufacturing, Jom 69(3) (2017) 472-478.
[21] M. Liu, S. Liu, W. Chen, C. Chen, Y. Lv, X. Zhang, P. Lei, Y. Lin, K. Zhou, Effect of trace lanthanum
hexaboride on the phase, grain structure, and texture of electron beam melted Ti-6Al-4V, Additive Manufacturing
30 (2019) 100873.
[22] S. Mereddy, M.J. Bermingham, D.H. StJohn, M.S. Dargusch, Grain refinement of wire arc additively
manufactured titanium by the addition of silicon, Journal of Alloys and Compounds 695 (2017) 2097-2103.
[23] M.J. Bermingham, S.D. McDonald, D.H. StJohn, M.S. Dargusch, Beryllium as a grain refiner in titanium
alloys, Journal of Alloys and Compounds 481(1) (2009) L20-L23.
[24] H. Okamoto, Desk handbook: phase diagrams for binary alloys, ASM international2000.
[25] M. Bermingham, S. McDonald, K. Nogita, D.S. John, M. Dargusch, Effects of boron on microstructure in cast
titanium alloys, Scripta Materialia 59(5) (2008) 538-541.
[26] S.C. Tjong, Y.-W. Mai, Processing-structure-property aspects of particulate- and whisker-reinforced titanium
matrix composites, Composites Science and Technology 68(3) (2008) 583-601.
[27] E. Fereiduni, A. Ghasemi, M. Elbestawi, Characterization of composite powder feedstock from powder bed
fusion additive manufacturing perspective, Materials 12(22) (2019) 3673.
[28] M.A. Balbaa, A. Ghasemi, E. Fereiduni, M.A. Elbestawi, S.D. Jadhav, J.P. Kruth, Role of powder particle size
on laser powder bed fusion processability of AlSi10mg alloy, Additive Manufacturing 37 (2021) 101630.
[29] U. Scipioni Bertoli, A.J. Wolfer, M.J. Matthews, J.-P.R. Delplanque, J.M. Schoenung, On the limitations of
Volumetric Energy Density as a design parameter for Selective Laser Melting, Materials & Design 113 (2017) 331340.
[30] ASTM, G99‐ 95a (2000) e1, Standard Test Method for Wear Testing with a Pin‐ on‐ Disk Apparatus, (2000).
[31] D. Odabas, Effects of load and speed on wear rate of abrasive wear for 2014 Al alloy, IOP Conference Series:
Materials Science and Engineering, IOP Publishing, 2018, p. 012008.
[32] A. Grabowski, M. Nowak, J. Śleziona, Optical and conductive properties of AlSi-alloy/SiCp composites:
application in modelling CO2 laser processing of composites, Optics and Lasers in Engineering 43(2) (2005) 233246.
[33] W. Zhou, X. Sun, K. Kikuchi, N. Nomura, K. Yoshimi, A. Kawasaki, Carbon nanotubes as a unique agent to
fabricate nanoceramic/metal composite powders for additive manufacturing, Materials & Design 137 (2018) 276285.
[34] E. Fereiduni, A. Ghasemi, M. Elbestawi, S. Dinkar Jadhav, K. Vanmeensel, Laser powder bed fusion
processability of Ti-6Al-4V powder decorated by B4C particles, Materials Letters 296 (2021) 129923.
[35] W. Liu, C. Chen, S. Shuai, R. Zhao, L. Liu, X. Wang, T. Hu, W. Xuan, C. Li, J. Yu, J. Wang, Z. Ren, Study of
pore defect and mechanical properties in selective laser melted Ti6Al4V alloy based on X-ray computed
tomography, Materials Science and Engineering: A 797 (2020) 139981.
[36] H. Salem, L.N. Carter, M.M. Attallah, H.G. Salem, Influence of processing parameters on internal porosity and
types of defects formed in Ti6Al4V lattice structure fabricated by selective laser melting, Materials Science and
Engineering: A 767 (2019) 138387.
[37] J. Lin, Y. Lv, D. Guo, X. Wu, Z. Li, C. Liu, B. Guo, G. Xu, B. Xu, Enhanced strength and ductility in thin Ti6Al-4V alloy components by alternating the thermal cycle strategy during plasma arc additive manufacturing,
Materials Science and Engineering: A 759 (2019) 288-297.
[38] Z. Zhao, J. Chen, X. Lu, H. Tan, X. Lin, W. Huang, Formation mechanism of the α variant and its influence on
the tensile properties of laser solid formed Ti-6Al-4V titanium alloy, Materials Science and Engineering: A 691
(2017) 16-24.
[39] J. Yang, H. Yu, J. Yin, M. Gao, Z. Wang, X. Zeng, Formation and control of martensite in Ti-6Al-4V alloy
produced by selective laser melting, Materials & Design 108 (2016) 308-318.
[40] E. Fereiduni, A. Ghasemi, M. Elbestawi, Selective laser melting of hybrid ex-situ/in-situ reinforced titanium
matrix composites: Laser/powder interaction, reinforcement formation mechanism, and non-equilibrium
microstructural evolutions, Materials & Design 184 (2019) 108185.
[41] S. Cao, Z. Chen, C.V.S. Lim, K. Yang, Q. Jia, T. Jarvis, D. Tomus, X. Wu, Defect, Microstructure, and
Mechanical Property of Ti-6Al-4V Alloy Fabricated by High-Power Selective Laser Melting, JOM 69(12) (2017)
2684-2692.
[42] M. Neikter, A. Huang, X. Wu, Microstructural characterization of binary microstructure pattern in selective
laser-melted Ti-6Al-4V, The International Journal of Advanced Manufacturing Technology 104(1) (2019) 13811391.
[43] M.J. Bermingham, L. Nicastro, D. Kent, Y. Chen, M.S. Dargusch, Optimising the mechanical properties of Ti6Al-4V components produced by wire + arc additive manufacturing with post-process heat treatments, Journal of
Alloys and Compounds 753 (2018) 247-255.
[44] B. Baufeld, O.V.d. Biest, R. Gault, Additive manufacturing of Ti–6Al–4V components by shaped metal
deposition: Microstructure and mechanical properties, Materials & Design 31 (2010) S106-S111.
[45] M. Narvan, A. Ghasemi, E. Fereiduni, M. Elbestawi, Laser powder bed fusion of functionally graded bimaterials: Role of VC on functionalizing AISI H13 tool steel, Materials & Design (2021) 109503.
[46] H. Shipley, D. McDonnell, M. Culleton, R. Coull, R. Lupoi, G. O'Donnell, D. Trimble, Optimisation of process
parameters to address fundamental challenges during selective laser melting of Ti-6Al-4V: A review, International
Journal of Machine Tools and Manufacture 128 (2018) 1-20.
[47] C.Y. Yap, C.K. Chua, Z.L. Dong, Z.H. Liu, D.Q. Zhang, L.E. Loh, S.L. Sing, Review of selective laser
melting: Materials and applications, Applied physics reviews 2(4) (2015) 041101.
[48] W.E. King, H.D. Barth, V.M. Castillo, G.F. Gallegos, J.W. Gibbs, D.E. Hahn, C. Kamath, A.M. Rubenchik,
Observation of keyhole-mode laser melting in laser powder-bed fusion additive manufacturing, Journal of Materials
Processing Technology 214(12) (2014) 2915-2925.
[49] T. Qi, H. Zhu, H. Zhang, J. Yin, L. Ke, X. Zeng, Selective laser melting of Al7050 powder: Melting mode
transition and comparison of the characteristics between the keyhole and conduction mode, Materials & Design 135
(2017) 257-266.
[50] C. Teng, D. Pal, H. Gong, K. Zeng, K. Briggs, N. Patil, B. Stucker, A review of defect modeling in laser
material processing, Additive Manufacturing 14 (2017) 137-147.
[51] D. Hann, J. Iammi, J. Folkes, A simple methodology for predicting laser-weld properties from material and
laser parameters, Journal of Physics D: Applied Physics 44(44) (2011) 445401.
[52] J. Yang, J. Han, H. Yu, J. Yin, M. Gao, Z. Wang, X. Zeng, Role of molten pool mode on formability,
microstructure and mechanical properties of selective laser melted Ti-6Al-4V alloy, Materials & Design 110 (2016)
558-570.
[53] T. Wang, Y. Zhu, S. Zhang, H. Tang, H. Wang, Grain morphology evolution behavior of titanium alloy
components during laser melting deposition additive manufacturing, Journal of Alloys and Compounds 632 (2015)
505-513.
[54] M.J. Bermingham, D.H. StJohn, J. Krynen, S. Tedman-Jones, M.S. Dargusch, Promoting the columnar to
equiaxed transition and grain refinement of titanium alloys during additive manufacturing, Acta Materialia 168
(2019) 261-274.
[55] F. Wang, S. Williams, P. Colegrove, A.A. Antonysamy, Microstructure and mechanical properties of wire and
arc additive manufactured Ti-6Al-4V, Metallurgical and materials transactions A 44(2) (2013) 968-977.
[56] I. Solodkyi, I. Bogomol, P. Loboda, D. Batalu, A. Vlaicu, P. Badica, Floating zone partial re-melting of B4C
infiltrated with molten Si, Ceramics International 43(17) (2017) 14718-14725.
[57] S.H. Whang, Rapidly solidified titanium alloys for high-temperature applications, Journal of materials science
21(7) (1986) 2224-2238.
[58] S. Tamirisakandala, R. Bhat, D. Miracle, S. Boddapati, R. Bordia, R. Vanover, V. Vasudevan, Effect of boron
on the beta transus of Ti–6Al–4V alloy, Scripta Materialia 53(2) (2005) 217-222.
[59] M.J. Bermingham, D. Kent, H. Zhan, D.H. StJohn, M.S. Dargusch, Controlling the microstructure and
properties of wire arc additive manufactured Ti–6Al–4V with trace boron additions, Acta Materialia 91 (2015) 289303.
[60] S. Sahay, K. Ravichandran, R. Atri, B. Chen, J. Rubin, Evolution of microstructure and phases in in situ
processed Ti–TiB composites containing high volume fractions of TiB whiskers, Journal of Materials Research
14(11) (1999) 4214-4223.
[61] A. Xue, X. Lin, L. Wang, J. Wang, W. Huang, Influence of trace boron addition on microstructure, tensile
properties and their anisotropy of Ti6Al4V fabricated by laser directed energy deposition, Materials & Design 181
(2019) 107943.
[62] R.R. Rashid, S. Palanisamy, H. Attar, M. Bermingham, M. Dargusch, Metallurgical features of direct laserdeposited Ti6Al4V with trace boron, Journal of Manufacturing Processes 35 (2018) 651-656.
[63] N. Hansen, Hall–Petch relation and boundary strengthening, Scripta Materialia 51(8) (2004) 801-806.
[64] J. Wyrzykowski, M. Grabski, The Hall–Petch relation in aluminium and its dependence on the grain boundary
structure, Philosophical Magazine A 53(4) (1986) 505-520.
[65] M. Galindo-Fernandez, K. Mumtaz, P. Rivera-Diaz-del-Castillo, E. Galindo-Nava, H. Ghadbeigi, A
microstructure sensitive model for deformation of Ti-6Al-4V describing Cast-and-Wrought and Additive
Manufacturing morphologies, Materials & Design 160 (2018) 350-362.
[66] X. Guo, L. Wang, M. Wang, J. Qin, D. Zhang, W. Lu, Effects of degree of deformation on the microstructure,
mechanical properties and texture of hybrid-reinforced titanium matrix composites, Acta Materialia 60(6) (2012)
2656-2667.
[67] H. Fukuda, T.-W. Chou, A probabilistic theory of the strength of short-fibre composites with variable fibre
length and orientation, Journal of materials science 17(4) (1982) 1003-1011.
[68] Y. Hu, W. Cong, X. Wang, Y. Li, F. Ning, H. Wang, Laser deposition-additive manufacturing of TiB-Ti
composites with novel three-dimensional quasi-continuous network microstructure: effects on strengthening and
toughening, Composites Part B: Engineering 133 (2018) 91-100.
[69] E. Orowan, Zur Kristallplastizität. III, Zeitschrift für Physik 89(9) (1934) 634-659.
[70] R. George, K. Kashyap, R. Rahul, S. Yamdagni, Strengthening in carbon nanotube/aluminium (CNT/Al)
composites, Scripta Materialia 53(10) (2005) 1159-1163.
[71] G.E. Dieter, D.J. Bacon, Mechanical metallurgy, McGraw-hill New York1986.
[72] M.A. Meyers, K.K. Chawla, Mechanical behavior of materials, Cambridge university press2008.
[73] L. Brown, W. Stobbs, The work-hardening of copper-silica v. equilibrium plastic relaxation by secondary
dislocations, Philosophical Magazine 34(3) (1976) 351-372.
[74] S. Schmauder, C. Kohler, Atomistic simulations of solid solution strengthening of α-iron, Computational
Materials Science 50(4) (2011) 1238-1243.
[75] P. Luo, D. McDonald, W. Xu, S. Palanisamy, M. Dargusch, K. Xia, A modified Hall–Petch relationship in
ultrafine-grained titanium recycled from chips by equal channel angular pressing, Scripta Materialia 66(10) (2012)
785-788.
[76] N. Ramakrishnan, An analytical study on strengthening of particulate reinforced metal matrix composites, Acta
Materialia 44(1) (1996) 69-77.
[77] K.D. Rekedal, Investigation of the high-cycle fatigue life of selective laser melted and hot isostatically pressed
Ti-6Al-4V, (2015).
[78] A. Muiruri, M. Maringa, W. du Preez, Evaluation of Dislocation Densities in Various Microstructures of
Additively Manufactured Ti6Al4V (Eli) by the Method of X-ray Diffraction, Materials 13(23) (2020) 5355.
[79] W.D. Callister, D.G. Rethwisch, Materials science and engineering, John wiley & sons NY2011.
Declaration of Interest Statement
Declaration of Interest Statement
The authors confirm no conflict of interest.